Related Experiment Video
Updated: May 22, 2025
The Mesenteric Lymph Duct Cannulated Rat Model: Application to the Assessment of Intestinal Lymphatic Drug Transport
Published on: March 6, 2015
Predicting lymphatic transport potential using graph transformer based on limited historical data from in vivo
Yunfeng Li1, Ruiya Liu2, Zonghao Ji3
1School of Pharmacy, China Pharmaceutical University, 24 Tongjia Lane, Nanjing 210009, Jiangsu Province, China; School of Publishing, Beijing Institute of Graphic Communication, 1 Xinghua Avenue (Band Two), Daxing, Beijing 102600, China; CNPIEC Kexin Digital Technology (Beijing) Co., Ltd, 16 Gongti East Road, Chaoyang District, Beijing, China.
This study developed AI models to predict lymphatic drug transport, overcoming challenges in drug delivery to the lymphatic system. The models correlate drug chemical structures with in vivo data, improving prediction accuracy for lymphotropic drug design.
Area of Science:
- Pharmacokinetics and Drug Delivery
- Computational Chemistry and Cheminformatics
- Artificial Intelligence in Drug Discovery
Background:
- The lymphatic system is a crucial therapeutic target for various diseases, but achieving effective drug delivery remains a significant challenge.
- Current approaches to designing lymphotropic drugs are often empirical, lacking clear mechanistic understanding and design criteria.
- Limited data accumulation due to complex experimental methods hinders progress in lymphatic drug transport research.
Purpose of the Study:
- To review and analyze existing data on lymphatic drug transport for 185 drugs.
- To develop and validate Artificial Intelligence (AI) models for predicting oral drug transport to the lymphatics.
- To investigate the relationship between drug physicochemical properties and lymphatic transport extent.
Main Methods:
- Systematic review and analysis of published lymphatic drug transport studies (direct and indirect measurements).
- Development of AI models: Graph Convolutional Network (GCN), Graph Attention Network (GAT), and Graph Transformer (GT).
- Utilized data augmentation and correlated in vivo data with drug chemical structures (Simplified Molecular Input Line Entry System - SMILES).
Main Results:
- The developed AI models successfully predicted lymphatic drug transport potential by linking in vivo data with drug chemical structures.
- Analysis revealed that lymphatic transport capability is not determined by a single physicochemical parameter (e.g., LogP, LogD7.4, molecular weight).
- The study provides a data-driven approach to enhance understanding and prediction of lymphatic drug transport.
Conclusions:
- AI-driven prediction offers a powerful tool to guide the design of lymphotropic drug candidates, moving beyond trial-and-error methods.
- A multi-parameter approach is necessary for understanding and optimizing drug transport to the lymphatic system.
- This research facilitates the development of novel therapeutics targeting diseases treatable via the lymphatic system.
